BY R. GREIG SMITH. 339 



of the slime by heating the emulsion under pressure in the 

 autoclave, a method which had been very successful in the 

 separation of the constituents of the arabin bacterial slimes. The 

 attempt failed, and no separation of the gum could be induced 

 by the method. 



In view of the probable impossibility of obtaining a clear solu- 

 tion of the gummy constituent, the whole slime was repeatedly 

 coacfulated with alcohol until most of the salts had been removed 

 and the bulk of the slime remained suspended in the dilute 

 alcohol as an opalescent solution. Saline flocculating agents were 

 then added. First potassium chloride threw down a fraction, 

 then strong alcohol precipitated a second fraction, barium chloride 

 flocculated a third portion. The mother liquor was now clear 

 and bright, but on boiling off" the alcohol a fourth fraction settled 

 out. All these fractions, with the exception of the last, formed 

 emulsions with water; the last fraction was more of the nature 

 of a suspension. The emulsions and the suspension behaved to 

 reagents in a manner precisely similar to the original slime and 

 to the residue which was not "milked" by the dilute alcohol. 

 From the similar behaviour of the fractions it was evident that 

 the slime contained but one gum constituent. Coagulation of 

 the emulsions was effected by alcohol, neutral and basic lead 

 acetates, milk of lime and baryta water. These reactions were 

 constant with the slime from the several races of the bacterium. 

 Coagulation was also effected by other reagents, but the reactions 

 could not be depended upon even with slimes from the same race. 

 For example, slime grown at 37^ gave a precipitate with 1 % 

 and 10 % copper sulphate, while when grown at 18*^ no precipi- 

 tate was obtained. With the crude slime a precipitate was 

 obtained with ferric chloride, but the partly purified slime gave 

 no precipitate. The ready solubility of the slime carbohydrate 

 in dilute acids may account for the irregular behaviour with the 

 salts of the metals. 



Of more importance than the reactions of the slime is the 

 nature of the essential carbohydrate. From saccharose, bacteria 

 can produce dextran,levan,galactan,arabinan-galactan and deriva- 



